Association between thyroid dysfunction and venous thromboembolism in the elderly: a prospective cohort study.

Background Venous thromboembolism (VTE) and subclinical thyroid dysfunction (SCTD) are both common in elderly patients. SCTD has been related to a hypercoagulable state and an increased thromboembolic risk. However, prospective data on the relationship between SCTD and VTE are lacking. Objectives To investigate the relationship between SCTD and recurrent VTE (rVTE), all-cause mortality, and thrombophilic biomarkers. Patients Elderly patients with VTE were studied. Methods In a prospective multicenter cohort, thyroid hormones and thrombophilic biomarkers were measured 1 year after acute VTE, as both may be influenced by acute thrombosis. We defined subclinical hypothyroidism (SHypo) as elevated thyroid-stimulating hormone (TSH) levels (4.50-19.99 mIU L-1), and subclinical hyperthyroidism (SHyper) as TSH levels of < 0.45 mIU L-1, both with normal free thyroxine levels. Outcomes were incidence of rVTE and overall mortality during follow-up starting after the 1-year blood sampling. Results Of 561 participants (58% with anticoagulation), 6% had SHypo and 5% had SHyper. After 20.8 months of mean follow-up, 9% developed rVTE and 10% died. The rVTE incidence rate was 7.2 (95% confidence interval [CI] 2.7-19.2) per 100 patient-years in SHypo participants, 0.0 (95% CI 0.0-7.6) in SHyper participants, and 5.9 (95% CI 4.4-7.8) in euthyroid participants. In multivariate analyses, the sub-hazard ratio for rVTE was 0.00 (95% CI 0.00-0.58) in SHyper participants and 1.50 (95% CI 0.52-4.34) in SHypo participants as compared with euthyroid participants, without increased levels of thrombophilic biomarkers. SHyper (hazard ratio [HR] 0.80, 95% CI 0.23-2.81) and SHypo (HR 0.99, 95% CI 0.30-3.29) were not associated with mortality. Conclusion In elderly patients, SHyper may be associated with lower rVTE risks. SHypo showed a non-statistically significant pattern of an association with rVTE, without increased mortality or differences in thrombophilic biomarkers.